The first published observations of this clinical entity had to wait until the early 1980s, after which medical and surgical intensive care units (ICUs) were well established. Once patients with sepsis and multiorgan failure were admitted to medical ICUs—and survived—they were there for the long run. Similarly, major surgeries in patients admitted to surgical ICUs could become more complicated. These patients became ventilator and dialysis dependent and had numerous infectious and thrombotic complications associated with prolonged bed rest. Their fragile state was obvious and there was trepidation about outcome and quality of life once they were weaned off the ventilator and intravenous infusions. It would not cross the intensivist’s mind that these patients who had survived this onslaught could have a neurologic disorder, and most attributed their frailty and poor mobilization to extreme deconditioning from a drawn-out illness.

Someone at some point consulted a neurologist with the specific question of whether there was a neuromuscular respiratory failure that would prevent the patient from coming off the ventilator. There might have been additional complaints of the patient, such as numbness, tingling, and burning of the hands and feet. In most, there was profound muscle weakness and atrophy that seemed out of proportion to prolonged illness, prolonged use of neuromuscular junction blockers (at the time), or a combination of the two. Someone at some point decided to do an electrophysiologic study of nerve conduction and muscle activity and found that the severe weakness, atrophy, and debilitating state had a neurologic basis. The original question of whether there was neuromuscular respiratory weakness from presumed phrenic nerve involvement was never fully satisfactorily answered. We now know the diaphragm looses its strength quickly in ventilated patients.

There were earlier sporadic publications of polyneuropathy after coma but after barbiturate intoxication. In 1941, Brunnschweiler and Caspers [1] found prolonged muscle weakness in patients after being in a coma for weeks, and Mertens and colleagues [2] two decades later were able to investigate a peripheral nerve in such a patient with fatal outcome and found neuronal edema, lymphocyte infiltrates and macrophages, and diffuse distribution of muscle involvement. Patterns of distribution of weakness in their cases are shown in Fig. 1 [2].

Mertens patients with diffuse muscle weakness (used with permission) From reference 2

In 1983, two abstracts by Charles Bolton and by Robert Roelofs [3, 4] were shown at the Annual Meeting of the American Academy of Neurology. Both authors described an axonal polyneuropathy, and Roelofs quickly suggested it may represent a new syndrome. Their patients had sepsis and multiorgan failure and as a result had developed profound weakness and prolonged respiratory failure [3, 4].

In Bolton’s series, five patients were seen in a 4-year time span. There were probably many more patients in the ICU, but he saw these patients because he was asked to do an electrophysiologic study focused on the axial muscles and muscles of breathing. In all patients, he noted that the polyneuropathy developed quite soon after admission to the ICU and likely at the onset of infection. The electromyography findings included fibrillation potentials, positive sharp waves, normal or mildly prolong motor nerve conduction velocities, and muscle biopsy samples showed mostly type II fiber atrophy. The clinical setting was uniformly sepsis often in association with severe pulmonary disease, which included acute pulmonary edema, empyema, and bilateral lobar pneumonia. Systemic illness always evolved into a multiorgan dysfunction. Cerebrospinal fluid was unremarkable in most patients and excluded an infection related Guillain-Barré syndrome, although the clinical and electrophysiologic pattern was axonal and not demyelinating. This was also supported by demonstrating an axonal injury rather than demyelination on a sural nerve biopsy sample. Both clinical reports speculated on the cause and logically included consideration of nutritional deficiencies, toxicity from antibiotics, or toxins associated with a bacterial infection. Naturally, it could not be determined whether the peripheral nerves were just part of the many tissues and organs directly damaged by sepsis or if the treatment of sepsis could be implicated. Pathology of the patients who died showed normal cranial nerves, but there was moderate to severe loss of large and small myelinated fibers in the median and lateral cutaneous, sciatic, common peroneal, superficial, and deep peroneal and sural nerves. Macrophages containing lipid debris were found but also occasional clusters of mononuclear inflammatory cells. Outcomes in survivors were good, but recovery was slow (Fig. 2). The abstract by Roelofs and associates [4] included four patients who were weaned off the ventilator after several months and electrodiagnostic studies showed active denervation characteristics. No explanation for the respiratory failure was provided [4].

Outcome of Bolton’s patients (used with permission). TPN, total parenteral nutrition

The term “critical illness polyneuropathy” was proposed, linking it to stay in the ICU and not to a specific illness. Several years later, in 1985, Op de Coul et al. [5] published a similar article on neuromuscular complications in patients given pancuronium bromide during mechanical ventilation. This was followed by a larger study by Zochodne and colleagues [6] from Bolton’s group who saw 5 patients in a 4-year period and then 14 patients in the next 2 years, suggesting increased recognition of the disorder. Shortly thereafter, cases of “acute severe muscle necrosis” were described by others [7]. Such unexpected muscle necrosis was tentatively associated with sepsis and high doses of corticosteroids or because of severe rhabdomyolysis.