Need for refined rapid kits, better staff training for accurate results
The technology behind antibody tests requires knowledge of viral protein against which the human immune system produces antibodies that flag or kill the virus.
Inaccuracy in results from China-made rapid kits led India to suspend antibody testing for coronavirus disease (Covid-19) last week. On Monday, the Indian Council of Medical Research asked for states to return the kits, citing problems with them highlighted in its own assessment.
Rapid tests do not assess current infection but are used to detect antibodies produced by the body to detect and neutralise Sars-CoV-2, the virus that causes Covid-19.
Unlike reverse transcription polymerase chain reaction (RT-PCR) tests that detect shards of the viral genetic material (RNA) to confirm current infection, antibody tests indicate past infection that may have been missed because it was mild or asymptomatic. Antibody tests are a crucial public health tool for surveillance, epidemic forecasting, and prevalence.
They can also (at least on paper) detect infections after around 10 days of a person being infected, although experts strongly recommend against their use for diagnosis.
“India so far uses basic lateral flow immunoassay (LFA) for antibody testing, which give yes and no results if the testing is done seven to 10 days after symptoms appearing. We don’t have data on how long Indians take to produce antibodies against Sars-CoV-2 ,” said Dr Shiv Kumar Sarin, director, Indian Institute of Liver and Biliary and Sciences, and the Chair of Delhi government’s five-member panel on coronavirus disease.
The development of immunity to an invading virus is a multi-step process that takes place over one to two weeks. The body responds to a viral infection immediately with a non-specific innate response in which macrophages, neutrophils, and dendritic cells slow infection, and this non-specific response is followed by an adaptive response, where the body makes immunoglobulin M (IgM) and Immunoglobulin G (IgG) antibodies that specifically bind to the virus to neutralise it.
The first wave of antibodies that fight infection are IgM, which are detectable around eight days after symptoms appear and indicate recent infection, and IgG bodies appear around three weeks after infection, indicating the person was infected and has since, recovered.
“We need kits that test for both IgM and IgG, so that both results can be combined . Detection rate is highest when IgM and IgG results are combined, but few vendors are making kits with strips for both types of antibodies,” said Dr Sarin, whose team was among the first to begin antibody testing at ILBS.
The technology behind antibody tests requires knowledge of viral protein against which the human immune system produces antibodies that flag or kill the virus. To make the kits, sections of the viral protein coat that triggers an immune reaction are produced in the lab using cell lines for inclusion in the serological test to detect antibodies.
First-generation antibody tests need to be refined and need at least six months to get the correct sensitivity and specificity to rule out false positives and false negatives. “Quality serological tests take around six months to be developed with ample data and validation. There’s a flood of tests with unproven quality, which cannot be depended on for diagnosis,” said Soumya Swaminathan, chief scientist, World Health Organisation.
Test sensitivity indicates the capacity to correctly diagnose those with the disease (true positive rate), while specificity indicates correctly identifying those without the disease (true negative rate). Studies show that positivity rises with time after symptoms appear, with rates of positive results rising after two weeks and peaking after 20 days of illness.
Training people when to do the antibody tests and how to read the results is critical for reliable diagnosis. “For accurate results, antibody tests must be done eight-10 days after the symptoms appear, and must be interpreted by highly-trained staff, with results with weak bands being reviewed by at least two to three trained people,” said Dr N K Ganguly, former director general, Indian Council of Medical Research (ICMR).
“International studies show IgM and IgG antibodies sometimes takes two to three weeks, which makes the IgM test less reliable than IgG test. Both should be combined for interpretation. The persons reading it should be well trained since a weak band needs to be interpreted correctly,” added Dr Ganguly.
Quality assessment done at ICMR-National Institute of Malaria Research in Delhi before the antibody kits manufactured by Guangzhou Wondfo and Zhuhai Livzon in China were sent to states found the efficacy, or capacity for correct diagnosis, to be 71% , against the declared sensitivity of 80% and specificity 84%. “This is a first-generation test developed in just three-and-a-half months and needs refinement, the variations cannot be ignored,” said Dr Raman R Gangakhedkar, chief, epidemiology and communicable diseases, ICMR.
“Indian scientists have the capacity to develop these tests, so the focus should now be on domestic development and manufacturing of serological assays and training people when to do tests and how to read results. They have good enzyme-linked immunosorbent assay (Elisa) and chemiluminescence-linked immunoassay (Clia) being used in the US; there’s nothing holding us back from making them,” said Dr Ganguly.
Interestingly, even in the US, the reliability of antibody tests have come under question. A test performance evaluation of 10 LFAs and two Elisa tests to detect Sars-CoV-2 antibodies showed heterogeneous assay performance, according to a yet to be peer-reviewed paper from multiple departments at University of California and Harvard Medical School, Boston. IgM detection was more variable than IgG, and detection was highest when IgM and IgG results were combined, with agreement between Elisas and LFAs results ranging from 75.8-94.8%, found researchers.
“There is no “gold standard” to identify true seropositive blood samples. The extent and time-course of antibody development are not fully understood as yet, and may vary between different populations, even among RT-PCR-confirmed cases... Our study also reinforces the need for assay validation using standardised sample sets,” researchers noted in the evaluation paper.
There is also an urgent need for good data from India. “Testing at the right time is critical but we still don’t have data on when antibodies appear in response to Sars-CoV-2. We assume its two weeks. It varies widely with infections -- IgM are detectable after 5-7 days in dengue, and 12 days for chicken pox. Until we have highly accurate tests, it’s better to err on the side of caution, as a false positive would mean a person who is not protected will believe he is. Serological tests are not meant for diagnosis,” said Dr Naveen Dang, consultant microbiologist and founder of Dr Dangs Lab.
To increase specificity, serological tests must select sections of the spike protein that are particularly distinct across coronaviruses, according to The Lancet. “For accuracy, you need both sensitivity and specificity. Lowering specificity increases the odds of cross-reactivity with other coronaviruses – there are seven that cause disease in humans -- so you may have false positives resulting from immunity to the four coronaviruses that cause mild cold-like symptoms,” said Dr Dang.