Quinine, originally made from the bark of the cinchona tree, is an alkaloid with a long history of medicinal use and has been available in synthetic form since the 1940s. It is perhaps best known for the treatment of malaria (e.g., Mandel et al.⁸) but is also prescribed for night cramps,⁹ with a dosage of 200–300 mg. In the past, it was used as an abortifacient.⁴ Acute quinine toxicity, or cinchonism, may be characterized by blindness, tinnitus, nausea, vomiting, cardiac dysrrhythmias, coma, and even death. Symptoms of cinchonism are likely with doses above 4 g, and as little as 8 g may be fatal.¹⁰ There is marked interindividual variation in susceptibility to quinine that may give rise to so-called idiosyncratic toxic reactions.^2,5,11 Serum quinine levels are a poor prognostic indicator.² In addition, this author has experience of one patient (unpublished data) who had been taking therapeutic doses of quinine for night cramps but who was a covert alcoholic and developed a typical retinopathy presumed consequent upon impaired liver function.

Acute visual disturbances occur in approximately 40% of patients, but less than a third of those suffer permanent visual impairment.² There is usually severe visual loss with gradual recovery over the subsequent days or weeks. These improvements in visual function, often dramatic, may incorrectly be ascribed to the effects of treatment. However, even though visual symptoms are common, patients usually present to physicians with a history of attempted suicide by quinine overdose, often in association with alcohol and/or other medication. This may delay referral to ophthalmologists and thence to electrophysiologists. Ocular quinine toxicity in humans has therefore been difficult to study in the acute phase.

In early presentation, fluorescein angiography shows attenuated retinal arteries with a return of normal choroidal fluorescence as the initial masking from acute retinal edema subsides. Histological examination may reveal collapse of retinal architecture, early gliosis, and vascular narrowing.^3,6 The pathological process in the chronic phase is due to either delayed or secondary ischemia, although clinically, the onset of vascular attenuation often heralds the return of central visual function. The acute effects are less well characterized; one group has suggested the possibility of retinal toxicity, being unable to find histological evidence of acute ischemic changes in an experimental model.³

There does not appear to have been an electrophysiological study of quinine retinopathy since the introduction of the International Society for Clinical Electrophysiology of Vision (ISCEV) standard to electroretinography. However, early studies report an electronegative response to a single bright white flash under dark adaptation, similar to the maximal mixed rod-cone response in the current ISCEV standard. There is preservation of the photoreceptor derived a-wave but marked reduction in the postreceptoral b-wave.^1,5,12 This appearance is superficially similar to that which occurs in association with ischemic damage to the inner nuclear layer consequent on central retinal artery occlusion. Typical findings appear in figure 75.1. In addition to the electronegative electroretinogram (ERG), note the marked delay and amplitude reduction in the 30-Hz flicker ERG. A further feature of note is the highly distinctive appearance when long-duration stimulation is used to assess ON and OFF pathway function. There is a profoundly electronegative ON response with virtually no b-wave, and an extended plateau to the OFF response d-wave, giving an overall waveform reminiscent of a sawtooth. This highly unusual waveform has been present in all cases of quinine retinopathy examined by the author in which ON and OFF response recording has been performed, but has not been recognized in other disorders and may be specific to quinine toxicity.

Figure 75.1.  

ERGs and PERGs in two patients with quinine retinopathy following overdose. Data are shown from the right eye. Patient A is a 72-year-old female who took an overdose of quinine as an abortifacient some 40 years prior to investigation. Visual acuity was 6/18. Patient B is a 55-year-old male with an 11-month history and visual acuity of 6/6. Both patients show very similar ERG findings. The rod-specific ERG is subnormal; there is an electronegative bright-flash ERG; cone flicker and single-flash ERGs are delayed and markedly subnormal; and ON-OFF response recording shows almost complete ON b-wave loss with an elevated and extended plateau to the d-wave. PERGs are undetectable despite the normal visual acuity in patient B.

In the acute phase, there is marked generalized retinal abnormality involving all ERG waveforms but not accompanied by the electronegative waveform, which only becomes apparent some weeks later (unpublished data). It is postulated that the initial disturbance reflects the known effects of quinine on cell membranes (e.g., Malchow et al.⁷). The acute effects, in which there may be reduction of vision down to no light perception, can then be ascribed to generalized retinal dysfunction involving all retinal cell types, including the retinal ganglion cells. The visual evoked potential (VEP) at this stage may be undetectable. As the acute effects resolve, visual acuity may recover, and the a-wave of the ERG may show recovery. The characteristic negative ERG waveform is then a feature. The VEP may show recovery, presumably reflecting recovery of ganglion cell function, but tends not to be of normal latency. The VEP delay can be assumed to be secondary to continuing macular dysfunction, as the pattern ERG at this stage may be undetectable despite normal visual acuity (see figure 75.1). It is uncertain whether the negative ERG reflects inner retinal damage consequent upon vascular spasm and vessel attenuation or whether the vessel narrowing reflects loss of demand from inner retinal structures.

To conclude, the presence of a profoundly electronegative ERG in a patient with marked field constriction, pale disks, and attenuated retinal vasculature raises the question of quinine toxicity. Directed questioning of the patient may be necessary to reveal the relevant history, usually of an overdose. The nature of the ON and OFF response abnormality seems characteristic.