Cold comfort in sickle cell disease

In this issue of Blood, Ivy et al¹ show that the alternative complement pathway drives vaso-occlusion and pain after cold exposure in a mouse model of sickle cell disease (SCD).

In Minnesota, they must think about cold a lot. Kofi Annan said that, as a young Ghanaian student in St. Paul, he found earmuffs “inelegant”—until it was −23°F. “I learned a precious lesson—that you don’t walk into a situation . . . and pretend you know better than the locals. . . . You better listen to them and look at what they do.”² Anecdotally, people with SCD have told providers that their symptoms are worse in the cold, in the heat, when the weather changes, and after getting out of the pool; epidemiological and pain testing suggests this as well.^3,4 In this work, Ivy et al have used a murine model of pain in SCD that listens to the locals—that is, may more closely mimic what people with SCD actually experience in their lives. The very interesting results are revealing about microvascular stasis, complement, and the systemic effects of cold exposure on pain and inflammation and may help us think about new approaches to sickle cell that extend beyond buttoning up one’s overcoat.

Sickle hemoglobin (HbS) is a gain-of-function mutation in the β chain of Hb, in which Hb molecules acquire the ability to polymerize when deoxygenated. This remarkable mutation has been described molecularly since at least 1949,⁵ but our knowledge about the relationship of this “simple” single base pair mutation of β-globin gene and its profound impact on quality of life for individuals living with SCD has been slow to come and is incomplete. Ivy et al’s work is a step in the direction of understanding this complicated relationship.

Based on earlier work,^6,7 this group was interested in the role of complement in SCD, and here, rather than using biochemical triggers of vaso-occlusive episodes (complement fragments, free Hb, hypoxia), they instead used exposure to cold (50°F). The Townes humanized mouse model of sickle Hb will, with time alone, develop hyperalgesia, and the investigators selected only non-hyperalgesic (low pain) mice for comprehensive studies of microvascular stasis in dorsal skin fold chambers, complement levels, quantitative pain thresholds, and hepatic inflammation. Cold exposure induced prolonged pain and vascular stasis in HbSS-containing mice; strikingly, this was associated with a marked increase in detectable plasma complement factor B's active subunit (Bb) and complement component 5a (C5a). (see figure). Pre–cold exposure treatment with anti-C5a and C5a receptor antibodies decreased levels of plasma C5a and also, somewhat mysteriously (as is the way with complement), Bb, while abrogating cold-induced vascular stasis and hyperalgesia in these mice.

A few additional points: these authors had earlier described endothelial upregulation of P-selectin (via Weibel-Palade bodies) after infusion of purified C5a. In the current, more physiological, studies, they showed that cold-induced microvascular stasis and pain were also inhibited by pretreatment with anti-P-selectin antibodies. Notably, P-selectin involvement in mechanical hyperalgesia was not seen in a different pain model using non-HbSS mice. This suggests that P-selectin selectively influences pain downstream of vaso-occlusion, rather than having a broad impact on pain sensation itself. Finally, the supplementary figures from Ivy et al showing nuclear factor κB phosphorylation and upregulation of vascular endothelial adhesion markers in the liver at least 4 hours after cold exposure remind us how systemic and long-lasting the aftereffects of temperature change may be in SCD.

This is a great thought-provoking study, and one that resists the clean arrows of mechanistic figures. So now, a paragraph of questions: What comes first following cold exposure: Hb polymerization and hemolysis, complement activation, vascular stasis, or pain?⁸ These authors showed that blocking complement affects vascular stasis and pain, but does primarily blocking Hb polymerization or hemolysis, vascular stasis, or pain affect the other phenotypes? Does cold exposure induce hemolysis? And how many circles are there within these circles—for instance, between inflammation and dysautonomia in HbSS, which may exacerbate aberrant responses to cold?⁹ This relationship is hinted at in supplementary figure 4 in Ivy et al, in which acetylcholine abrogates microvascular stasis following cold exposure in HbSS; it would be good to know whether complement levels or pain are lessened there as well. Also unclear is how these data apply to variant disease hemoglobin SC (HbSC) or hemoglobin SBeta+thalassemia (HbSβ⁺ thalassemia), which was not tested here.

The hyperalgesic HbSS mice are also very interesting. Without cold exposure, these mice have pain and elevated complement fragments, both of which are lessened by treatment with anti-complement antibodies. Nothing is reported yet about vascular stasis or response to cold in these mice. Can these mice tell us something about the pathophysiology of chronic pain in HbSS?

This work reminds us how profoundly the environment can affect people with SCD; it goes without saying that comfort and warmth benefit all chronically ill people, and in those with SCD, the latter may literally be true. However, as eager as one is to make meaningful change in symptom burden for people with SCD, the specter of preemptive and chronic complement blockade in patients who may have hyposplenism is daunting (NCT 05075824 and NCT 05565092¹⁰) and will need to be monitored very closely, so that infectious risks can be identified and minimized.

Conflict-of-interest disclusure: J.L. receives research support from Pfizer, Novo-Nordisk, and the American Society of Hematology. E.S. receives research support from CSL Behring and National Institutes of Health.