a free page

My friend Baldolino Calvino invited me to write a post on his blog lost in the far reaches of the net. Why did I decide to accept? To paraphrase Sabine Hossenfelder, “to keep my sanity”. Possibly I will need to write many more posts if given the opportunity to achieve this.

I begin by reporting that a professional colleague contacted me to make me a proposal. He works with a pattern recognition technique used in the medical diagnosis of cancer (if you thought of ‘pathology’, you might be right). In his field, he increasingly uses molecular biology markers to aid in diagnosis, which has helped to transform an often subjective practice into a more objective and easily reproducible model. Even so, in his own words, “the intra- and inter-observer variation is very large”. Specific diagnoses are associated with specific marker patterns, and this is verified by scientifically tested evidence. So, there is a “package” of markers used for each diagnostic entity, with higher or lower sensitivity or specificity for the diagnosis. One of the concerns in a country and in a region like the one where we live is the lack of resources to carry out all the tests needed to increase the certainty of a diagnosis. Thus, it is necessary to work with the cost-benefit ratio of the procedures and choose a minimum set of markers able to correlate with an acceptable diagnostic capacity. In other words: one works not in the ideal situation, but in the possible situation. Sometimes it is not possible to give certain diagnoses with certainty, but this is not always clinically relevant.

An example to illustrate: in the 2016 revision of the classification of central nervous system tumors by the World Health Organization, the group of Embryonal Tumors includes a series of diagnoses that need one or more molecular markers to be classified with assurance. The entity embryonal tumor with multilayered rosettes (ETMR) has two variants: one that carries amplifications of the cluster of genes C19MC, which encodes miRNA; and one that does not have such molecular changes. It is obvious that this specific molecular marker(s) is needed to differentiate the two variants. In our service, however, these markers are not available. Thus, patients with compatible histology and/or tumor topography (it is typically a midline tumor, most commonly in the third ventricle) end up being diagnosed with Embryonal Tumor not otherwise specified (NOS), with a note indicating the possibility of being ETMR. However, this has no clinical relevance, as all embryonal tumors of the central nervous system that are not medulloblastomas or atypical teratoid-rhabdoid tumors are treated in the same way. That is, an ETMR or an NOS embryonal tumor receive the same treatment and have a similar prognosis (within certain limits). Thus, investing in this specific marker becomes a secondary priority in a scenario of limited resources.

The proposal: he informed me that he has some markers that are “left over” from underused kits and that, close to expiration (or perhaps shortly after) he could use to mark some specimens. He cited 3 unrelated markers and confessed not knowing the frequency of their occurrence in certain diagnoses. And finally, he asked me if there was no way to “do something” to “take advantage of” this material. By do something he means taking a set of samples (whose clinical relevance I could select) to check for the presence of these markers and, perhaps, draw some relevant conclusion (I think that last part would be my responsibility). There are a number of very obvious problems (to me at least) with this proposal, and they are worth examining individually:

Waste of expensive material in oncology practice:

This is not a new issue, but it has had some impact recently, with possible ethical implications [1,2]. Here in our country, the rules are different, but they also encourage the waste of resources. The molecular exam kits referred to in the case are purchased for a minimum number of cases (usually on the order of one or more hundreds) and have a limited expiration date after opening for use. Thus, if there is not enough demand for a particular diagnostic service, a good part of the kit is lost, which makes the procedures costly, in addition to the obvious waste of resources. This also occurs with medicines, but we will stick to the case of supplies for molecular diagnostic tests, which are expensive.

The ideal and logical solution would be to concentrate the specimens in reference labs, so that the underutilization of the reagents would be minimal. Unfortunately, a range of obstacles exist for this. In this case specifically, the service that performs the exams is a company hired by the public service in which I work. Companies have an incentive to limit cooperation: competition. Although the cost savings associated with outsourcing services to centralized units ends up outweighing the competition, it still encourages companies to offer a service directly, rather than hiring a third party to do so. As the amount passed on to companies is enough to account for a certain degree of waste, it ends up “being paid”. A policy that explicitly reduced the resource payment associated with waste could force smaller companies to outsource infrequent services to others that centralize service to larger regions. This already occurs in some situations, in which it is more economical to send the specimen to another center than to perform the analysis in loco.

Public contracts in our country, as a rule, encourage a wasteful policy, accentuating the difficulty of offering more complete services due to limited resources. Several reasons contribute to this: little flexibility of contracts, reduced transparency, occurrence of corruption at various levels, or simple technical unpreparedness of those involved. Thus, the fact that expensive reagents are routinely left over and thrown away is a serious problem and their use for some scientific research does not seem to be the best solution.

Why do scientific work:

A blog that discloses an interesting open research notebook model focused on neuro-oncology features a post about why do clinical research. With good humor, he gathers “wrong reasons” to do clinical research. The proposal made to me by the colleague fits well into the fourth item: “I gather this huge amount of data. There must be something that can come out of it!”

All clinical scientific research comes (or should come) from a situation where we need to solve a problem and we still don’t know how. How to cure a certain disease? Between two effective drugs for a disease, which is better? What is the best way to confirm the diagnosis of a particular disease? Are there differences in the incidence of a disease between ethnic groups, sexes, education levels, economic profile, etc? Have a problem, turn it into an objective question, and there will almost always be a way to find the answer to it through a clinical trial design. However, this is not the main motivation for doing clinical research today. Most professionals who end up interested in clinical research (if we can call it that), at least in our country (but I suspect that also abroad) do not want to answer a question, but earn a title on their resume. Mostly, graduate students looking to “vitaminize” their curriculum to meet the standards of their graduate program or improve their chances of getting an academic position. In some cases, pure and simple vanity is the only motivation.

So it’s no surprise that a motive as quirky as “there’s a reagent left, let’s do something with it” becomes perfectly possible. Even so, if there was a rationale for a research, waiting for resources to be carried out and the reagents available in this way could supply at least partially a resource for such an initiative, this would turn a debatable idea into a good opportunity. Clearly, however, this is not the case with the proposal made to me. There is no rationale behind it. In fact, the colleague just wants me to find a rationale, in an inversion of the methodology that can lead to serious biases. Needless to say, from me there won’t be much resonance for this.

The tragedy of clinical research in our country:

In the last 2 years, the current government committed the [biggest cut][cut] of resources for Science and Technology in history. Not to mention the repeated cuts in education. Added to this is the current scenario of already reduced investments in S&T and education and the population’s deficient access to quality basic and higher education. Our people have levels of education below what would be expected of a nation that is currently going through the stage of predominance of the youngest age groups in the age pyramid. Historically, this is the stage where the workforce has the greatest potential and where investment in education reverts better future results. In other words, we are missing a historic window of opportunity in our country that will leave consequences for future generations.

From the point of view of clinical research in our country, it is practically non-existent. Not only that, but there is a widespread culture that it is secondary. Some time ago, I had the opportunity to witness a sentence triggered by a colleague during a meeting of my specialty in my service: “I don’t care about knowing molecular biology, I just care how to use it in practice”. In this way, a future is established in which the main role of our country will be that of technology consumer, in all areas, including the clinic, losing the ability to produce technology. It’s the sad tropics. In the meantime, there will always be graduate students wanting to do some “work” with “leftovers” of expensive resources wasted somewhere.

References:

1. Goldstein DA, Hirsch A. A Policy That Encourages Waste of Expensive Medications—The JW Modifier. JAMA Oncol. 2018;4(2):155–156. doi:10.1001/jamaoncol.2017.3997
2. Zapata JA, Lai AR, Moriates C. Is Excessive Resource Utilization an Adverse Event?. JAMA. 2017;317(8):849–850. doi:10.1001/jama.2017.0698