Introduction
I welcomed the arguably provocative title which I was allocated
for this seminar because it provides an opportunity to discuss
three potentially philosophical, but (in my view) important questions
- a) what differences are there, if any, between a ‘surgical
pathologist’ and an ‘ultrastructural pathologist’?
b) is it clinically relevant to spend time and effort in subclassifying
soft tissue tumors? c) can the cost of electron microscopy be
justified in a clinical setting (in the particular context of
soft tissue neoplasia)?
Types of Pathologist
In reality, the decision to initiate (or Utilize) electron microscopy
(EM) for clinical diagnosis is almost invariably taken by a surgical
pathologist - and the overwhelming majority of EM is interpreted
by surgical pathologists. The number of institutions which can
afford to employ separately designated ultrastructural pathologists
is extremely small -and, almost by definition, if such institutions
have decided that they value EM enough to hire such a specialized
individual, then it is highly unlikely that their goals (or views)
regarding EM will differ to any significant degree.
For the most part, surgical pathologists and ultrastructural pathologists are virtually indistinguishable-hence the implicit notion, expressed in the title of this talk, that surgical pathologists might discourage or deplore the use of electron microscopy in a clinical (i.e. practical) setting serves only to highlight our specialty’s notable propensity for self--immolation. In almost any other specialty in medicine, if there is a technique which can provide information which is not otherwise available and especially if such information might be of clinical utility then there is no way in which it will be discarded or scorned -and if it is (probably mistakenly) regarded as expensive then this would add only to its mystique, while prompting efforts to reduce running costs. It is perhaps a special feature of pathology, which to a significant degree is technique-based, that our pursuit of fashion often leads us to belittle methodologies which are no longer ‘new’. A separate subset of our colleagues (perhaps the ‘surgical pathologist’ to which the title refers) has a well-known tendency to regard the need to employ any ‘special’ technique (whether EM, immunohistochemistry or molecular genetics) as evidence of diagnostic feebleness or lack of machismo - imagine what would have happened to radiology if the same approach had been applied to CT or MRI imaging The frequent unwillingness of surgical pathologists to embrace new (and not so new) technologies (so long as they are first critically assessed) and the consequent failure to strengthen their diagnostic armamentarium (which can only enhance the value, as well as status, of our specialty) is a cause of concern. Perhaps we need to remind ourselves that many of today’s more senior chairmen made their names (and reputations) in electron microscopy when (between perhaps 1960 and 1975) it was one of the premier research tools in pathology!
The Value of Classification
In the specific context of this talk, the question arises as to
the clinical value of electron microscopy in diagnosing (or subclassifying)
soft tissue tumors. This is a long-standing controversy which
centers principally around the question of whether the time and
effort (and perhaps cost - discussed below) spent on examining
these tumors ultrastructurally adds anything useful to the patient’s
care. The principal arguments raised against EM are either a)
that neither treatment nor prognosis are likely to be altered
or b) that other (and perhaps ‘cheaper’) techniques,
such as immunohistochemistry, can more reliably achieve the same
goal.
To me, these arguments cut to the very heart of the raison d’être of diagnostic anatomic pathology - both with regard to the individual patient and to the study of disease(s) in general. It is only through the accrual of additional or more detailed information that any progress is made- the nihilists’ view that sarcoma patients will not be assisted by EM because presently available treatment options are limited is comparable to proposing, in the preantibiotic area, that subtyping of bacteria would be a waste of time. It is true tat many diagnoses nowadays can more readily be confirmed by immunohistochemistry than EM but to suggest that EM is no longer worth considering as a diagnostic modality would be both premature and short-sighted.
It is not difficult to provide a list of examples in which EM has played (and continues to play) a significant role and, even if this list is selective in the context of soft tissue tumors as a whole, that is not grounds for abandoning the technique. Some of these examples reflect the fact that immunohistochemistry also has limitations, either due to antigen non-specificity, phenotypic overlap, aberrant or very focal antigen expression among other reasons - so these techniques are complementary. In fact, aside from examples listed below, it is generally true to say that EM is often more objective than immunohistochemistry in the final analysis of any problematic soft tissue tumor. Specific individual examples include:
A broader and perhaps more important example of the utility of EM is in the subcategorization of pleomorphic (MFH-like) sarcomas. EM played an important role in the recognition that many of the lesions formerly lumped together as so-called MFH actually show a specific and definable line of differentiation. The clinical relevance of such subcategorization is now beginning to emerge, since (for example) it is increasingly clear that pleomorphic sarcomas showing myogenic differentiation have a much higher metastatic rate (and poorer outcome) than other types of pleomorphic sarcoma - this provides a more rational basis for therapeutic trials and perhaps for aggressive chemotherapy - and EM can help to identify these cases. Furthermore, since data suggest that adult rhabdomyosarcomas are much more often chemoresponsive than leiomyosarcomas (using presently available regimes) and since both may stain only for muscle actin and desmin, then EM may be the only way to make this distinction.
What about cost?
The issue of cost-benefit and electron microscopy is a long-standing
controversy, full discussion of which is beyond the scope of this
seminar. It is clear that the major (and some would argue overwhelming)
cost is capital equipment, particularly if one seeks to purchase
a new electron microscope. While it is true that electron microscopes
are often long-lived (frequently being used for 20-25 years),
the majority of hospitals are unwilling (or unable) to contemplate
the cost of replacement - a situation which now faces the numerous
hospitals who invested in this technology in the early l970s.
Nevertheless, major institutions (particularly cancer centers
and hospitals with a high volume of renal disease) continue to
bear this cost because (in such centers) the technique is generally
considered useful. If one sets aside this equipment cost, the
costs of laboratory consumables and technical labor are entirely
comparable to other specialized diagnostic techniques in pathology.
The problem more often is one of test volume (and hence ensuring
that one gets a decent return on EM technicians’ salaries):
increasingly there is a tendency to seek individuals who can serve
other functions (e.g. working in immunofluorescence, immunohistochemistry
or photography) if their time is not fully occupied by EM.
One potentially workable solution, which would address both sides of this coin, would be to make more use of specialist off-site or centralized EM laboratories which could be cost-effective through handling significant test volume. It should be straightforward for almost any pathology laboratory to take tissue from a specimen for EM (into suitable fixative) and then send it away for processing and examination if the need for EM becomes clear.
A separate aspect of the cost issue is that of making a significant charge to the patient (or their insurer) for a test that might not ultimately prove to have been clinically useful or which was arguably unnecessary. A partial but simple way around this is, having taken tissue into EM, not to run the ‘test’ until the need becomes clear from initial light microscopic examination. Conversely we should remember that by no means all clinical or laboratory investigations prove to be useful in the final analysis - but this does not preclude their potential use at the time that they are ordered!
Limitations
While personally I do not doubt the practical value of EM as an
adjunct to the diagnosis of soft tissue tumors, one cannot ignore
the limitations of this technique both in this specific context
and in general. In the specific context of soft tissue neoplasia,
the principal limitations at the present time seem to be (among
others) a) our inability to define a myofibroblast in any unanimous
way, b) our inability to reliably distinguish smooth muscle cells
from myofibroblasts, c) our difficulty in reliably distinguishing
true (non-membrane-bound) lipid droplets from lysosomes in lipoblast-poor
tumors, d) the morphologic overlap among fibroblastic tumors of
quite different biologic potential, and e) our difficulty in reliably
identifying poorly differentiated endothelial tumors. In terms
of more general limitations, applicable to diagnostic EM however
used, these are well known and include the problems of very limited
sample size, the fact that diagnostic features may be present
in a minority of cells, the fact that tissue fixation is frequently
suboptimal (because EM is thought of too late!) - and the regrettable
but increasing non-availability of this diagnostic technique!