Confounding Factors Affecting the National Institutes of Health (NIH) Chronic GVHD Organ-Specific Score and Global Severity

The 2005 NIH chronic GVHD organ and overall clinical severity score is based on the assessment of current status regardless of whether abnormalities are caused by chronic GVHD or other conditions. The score assignment does not require knowledge of past manifestations, attribution, or whether GVHD is still active; it simply describes the patient's current condition. The aim of this prospective study was to analyze confounding factors affecting organ-specific 0-3 scores and global severity scores in patients with chronic GVHD. Patients and Methods: All 139 patients with NIH chronic GVHD evaluated at our center between January and September 2013 were included in this study. Medical providers completed the NIH 0-3 organ-specific scoring form with two additional questions for each organ to identify abnormalities that were 1) not attributed to chronic GVHD (yes/no), or 2) attributed to chronic GVHD plus other causes (yes/no). Causes of abnormalities attributed to causes other than GVHD were described by the medical evaluator. Results: Sixty-nine (50%) of the 139 patients had abnormalities attributed only to chronic GVHD. Fifty-nine (42%) patients had abnormalities not attributed to chronic GVHD in at least 1 organ, and 24 (17%) had abnormalities attributed to chronic GVHD plus other causes in at least 1 organ. The distribution of organ involvement according to attribution of the abnormalities is shown in the Table. Abnormalities not attributed to chronic GVHD were observed most frequently in lungs, gastrointestinal tract and skin sites. The most common causes of abnormalities attributed to other causes than GVHD included: conditions prior to the transplant, sequela from prior GVHD, deconditioning, infections and medications. When scores for organs or sites with abnormalities attributed to other than chronic GVHD were rescored as zero, the global severity score shifted downward from mild (n=7), moderate (n=2), and severe (n=1) to none and from moderate (n=8) and severe (n=1) to mild (Figure).  Conclusion: A modest downward shift of the global severity score occurs when confounders are taken into account in the 0-3 organ-specific scoring system. Our findings support the need for clarification of the NIH diagnosis and staging consensus criteria.

Table.  Distribution of organ involvement according to attribution of abnormalities
	Normal N (%)	Abnormalities attributed only to GVHD N (%)	Abnormalities attributed to GVHD plus other causes N (%)	Abnormalities  attributed only to other causes N (%)
Skin	55 (40)	63 (45)	4 (3)	17 (12)
Mouth	70 (50)	64 (46)	4 (3)	1 (1)
Eyes	71 (51)	59 (42)	-	9 (7)
GI tract	110 (79)	15 (11)	3 (2)	11 (8)
Liver	99 (71)	27 (19)	8 (6)	5 (4)
Lungs	97 (70)	18 (13)	3 (2)	21 (15)
Joint	97 (70)	32 (23)	3 (2)	7 (5)
Genital	123 (89)	10 (7)	2 (1)	4 (3)

Figure: Distribution of global severity scores according to NIH chronic GVHD consensus criteria with and without confounders