There was a significantly higher number of patients with multiple oral lesions at the time of oral HIV-KS diagnosis than patients who had single lesions (). With decreasing order of frequency, the gingiva, hard palate, oropharynx (upper and lower retromolar area, and soft palate), alveolar mucosa, and the dorsum of the tongue (Figures , , , , and ) were the sites most commonly affected (), conforming to other reports in the literature [4
]. In none of the 37 patients included in this study was the floor of the mouth or the ventral/lateral surface of the tongue affected. It is unknown why HIV-KS has the tendency to affect only certain oral sites, but not others.
Figure 1 Exophytic oral HIV-KS lesions on the anterior maxillary and mandibular buccal gingiva of a 31-year-old male with a CD4+ T-cell count of 5cells/mm3.
Figure 2 Exophytic confluent oral HIV-KS lesion on the hard palate in a 31-year-old male patient with a CD4+ T-cell count of 5cells/mm3.
Figure 3 Exophytic oral HIV-KS lesion on the lower right retromolar area extending into the oropharynx in a 29-year-old female patient with a CD4+ T-cell count of 49cells/mm3. The patient died six weeks after her oral HIV-KS diagnosis.
Figure 4 Exophytic oral HIV-KS lesions on the alveolar and labial mucosa in a 54-year-old male with a CD4+ Tcell count of 258cells/mm3. The patient died 15 weeks after his oral HIV-KS diagnosis.
Figure 5 Macular/nodular lesion on the dorsum of the tongue in a 44-year-old female patient with a CD4+ T-cell count of 13cells/mm3. The patient died five weeks after the diagnosis of her oral HIV-KS.
Forty six percent (46%) of the patients with oral HIV-KS in this study cohort did not know their HIV-serostatus at the time of oral HIV-KS diagnosis, implying that oral KS in the Ga-Rankuwa area in South Africa may serve as an indicator of HIV infection. Although the prevalence of HHV8 infection in South Africa is relatively high [22
], there were no recorded cases of oral KS in HIV-seronegative subjects during the study period, suggesting that endemic African KS is not frequent in the Ga-Rankuwa area of South Africa.
Seven patients were on HAART at the time of oral HIV-KS diagnosis and although at this time their average CD4+ T-cell count (90
) was lower than the average CD4+ T-cell count of the HAART-naïve patients (114
) (), this difference was not statistically significant. This seemingly surprising finding that the patients on HAART had a lower CD4+ T-cell count than the HAART-naïve patients could be attributed to the fact that in South Africa, HIV-seropositive persons who rely on provincial (governmental) services for their medical care have to abide by an official policy that HAART can be provided only when their CD4+ T-cell count has dropped below 200
, and that some people who have medical conditions suggestive of HIV disease are reluctant to undergo serological testing for HIV and often prefer to be treated by traditional healers. As a result, HIV infection is often diagnosed and HAART is often introduced only late in the course of their HIV disease when the CD4+ T-cell counts have already fallen substantially below 200
As a consequence of HAART being introduced only when the CD4+ T-cell count is already very low, there will be a lower level of reconstitution of the CD4+ T-cell number compared to the level of reconstitution when HAART is started at a higher CD4+ T-cell count [8
]. In addition, in provincial (governmental) medical facilities in South Africa, as in other countries in sub-Saharan Africa, monitoring the effectiveness of HAART is not always as efficient as it is in developed countries due to limited resources [28
]. All these factors may explain why the average CD4+ T-cell count of our study cohort was low at the time of oral HIV-KS diagnosis, why the number of HAART-naïve patients was high, and why the CD4+ T-cell counts of patients on HAART was also low.
It has been reported that in 22% of HIV-seropositive subjects with KS, the initial presentation of HIV-KS is in the mouth, and that in up to 70% of subjects with HIV-KS, the mouth will sooner or later be affected [12
]. As our study was designed to include only patients with oral HIV-KS regardless of whether they had extraoral KS, our findings cannot be compared to the findings of other studies reported in the literature in which the inclusion criterion was that of patients having HIV-KS, but who may or may not have had oral HIV-KS. However, our data show that in 21 patients (57%) the initial presentation of HIV-KS was in the mouth, and that in six patients the initial presentation of HIV-KS occurred concurrently in the mouth and on the skin. Ten patients (27%) developed cutaneous KS before and three after oral HIV-KS diagnosis. Eighteen patients (49%) with oral HIV-KS did not develop cutaneous HIV-KS during the observation period. These findings emphasise that in many cases the mouth may be the only body site affected by HIV-KS.
Facial lymphoedema may occur before, at the time of, or after oral HIV-KS is diagnosed. Facial lymphoedema which develops in parallel with the progression of oral HIV-KS disease is an indicator of poor prognosis [7
] and foretokens death [30
]. In this study, the nine patients who had facial lymphedema had extensive exophytic oral lesions, severe immunosuppression, and average CD4+ T-cell count of 28
and all died very soon after the onset of facial lymphedema, regardless of HAART.
The pathogenic mechanisms that cause facial lymphoedema in association with oral HIV-KS are obscure. However, as oral KS lesions and oral fluids of HIV-seropositive subjects carry a high HHV8 load, and as advanced exophytic oral HIV-KS lesions have a higher HHV8 load than initial maculopapular lesions [31
], it is possible that in the presence of exophytic oral lesions, lymphatic obstruction secondary to HHV8-induced proliferation of endothelial cells, and/or compression of lymphatics by rapidly progressing oral HIV-KS, will bring about leakage of protein-rich fluid into the interstitial spaces, promoting the development of facial lymphoedema [32
]. Therefore, it is likely that treating exophytic oral HIV-KS lesions with cytotoxic chemotherapy may result in the shrinkage of oral lesions with the subsequent decrease in HHV8 load in the affected tissues, thus reducing the risk of developing facial lymphoedema. This possibility needs further investigation.
To the best of our knowledge, this is the first report in the literature documenting the prevalence of IRIS-associated oral HIV-KS in a population of patients with oral HIV-KS regardless of whether they had extraoral KS. One patient in this case series had IRIS-associated HIV-KS. This patient was successfully treated with systemic cytotoxic chemotherapy and surgical excision. A comprehensive description of the case report of this patient has been published previously [34
]. At the time of writing this paper, the patient was still alive 5.5 years after the treatment of IRIS-associated oral HIV-KS and currently her CD4+ T-cell count is 383
. This is in line with reports in the literature documenting that IRIS-associated oral HIV-KS responds well to conventional therapy [35
During the study observation period, 21 of the 37 patients (57%) died, on average within 13.6 weeks from the time of oral HIV-KS diagnosis; nine were lost to follow-up; 7 patients survived (average period of follow-up of 91 weeks). The average CD4+ T-cell counts of the patients who were alive at the end of the study observation period was 166
at oral HIV-KS diagnosis and the CD4+ T-cell counts of the patient who died was 64
, at oral HIV-KS diagnosis. The difference in these CD4+ T-cell counts was statistically significant (P
= 0.016). This suggests that a low CD4+ T-cell count at the time of oral HIV-KS diagnosis is a strong indicator of poor prognosis. Unfortunately, we were unable to determine the cause of death of our patients and therefore it is unknown whether they died as a direct consequence of their HIV-KS disease.
Of those who survived, in three patients oral HIV-KS completely resolved following various treatment modalities. One of these patients was treated with HAART in combination with systemic cytotoxic chemotherapy and surgery, one with HAART in combination with cytotoxic chemotherapy, and one patient with HAART and surgery. In four patients, the oral HIV-KS remained unchanged or shrunk. These patients were treated with HAART or with HAART in combination with local cytotoxic chemotherapy.
HAART is used to treat HIV infection. Even though effective HAART does not directly influence HHV8 replication, by reducing HIV load with subsequent improvement in the host-immune status, it indirectly brings about a decrease in the incidence and prevalence of HIV-KS and an improvement in the clinical manifestation of pre-existing HIV-KS disease [29
]. HAART may also have a direct anti-angiogenic effect on KS. However, HAART does not ensure that HIV-KS will not develop, and despite HAART, KS remains the most frequent HIV-associated neoplasm [30
In this study, seven patients have been on HAART at least four weeks before their oral HIV-KS diagnosis, confirming that HIV-seropositive subjects on HAART may develop KS. Eight of the 21 patients who died during the observation period were on HAART as a sole modality of treatment. These patients died on average 4.4 weeks after their oral HIV-KS diagnosis. During this period, they experienced worsening of their oral HIV-KS disease. This suggests that although introduction of HAART should be the first line of therapy for HAART-naïve HIV-seropositive subjects with oral KS, HAART by itself may not be effective in controlling oral HIV-KS disease. Two of the 21 patients who died were concurrently treated with HAART and with local cytotoxic chemotherapy.
Eleven of the 21 patients (52%) who died during the observation period received neither HAART nor any other treatment for their oral HIV-KS. The average time from oral HIV-KS diagnosis to their death was 20 weeks. The paradoxical finding that in this study HAART-naïve patients lived longer than patients on HAART might be attributed to skewed statistics associated with the small number of patients, or to ineffective HAART that was started late in the course of oral HIV-KS disease when the CD4+ T-cell count had already fallen very low.
The small number of patients who received treatment for HIV-KS in this study prevents drawing conclusions regarding what is the best treatment approach to control the progression of oral HIV-KS and to improve the prognosis of the patients. However, as reported elsewhere [37
], it seems that exophytic oral HIV-KS lesions are best treated with HAART and systemic cytotoxic chemotherapy, and once the lesions have shrunk and become surgically accessible, they should be excised.
In developed countries, HIV-KS predominantly affects males. However, in many resource poor countries in sub-Saharan Africa where HIV infection is endemic and young females aged 15–24 years are more frequently infected with HIV than males, there is almost an identical incidence of HIV-KS in males and females [19
], and at the time of HIV-KS diagnosis females present with a more advanced disease than males [19
]. In our study, more females than males were affected by oral HIV-KS and although not statistically significant, females had a lower average CD4+ T-cell count (85
) than males (141
) at the time of oral HIV-KS diagnosis, in line with other studies documenting that females with HIV-KS have more severe immunodeficiency than males with HIV-KS [39
]. However, in contrast to other studies reporting that females are younger than males at the time of HIV-KS diagnosis [39
], the age of the females (33 years) and males (34 years) in our cohort was very similar.
In spite the fact that the CD4+ T-cell counts of females were lower than males at oral HIV-KS diagnosis, the differences between the percentage of males and females who survived or died were not statistically significant. Other reports in the literature also note that gender differences do not influence survival of patients with HIV-KS [19
Two patients were children. Both were boys, aged 10 and 11 years. This supports data reported from other areas of sub-Saharan Africa, that children are not uncommonly affected by HIV-KS. In fact, in Zimbabwe, the most frequent malignancy in children between 1 and 14 years of age is HIV-KS [38
]. This is probably owing to the high prevalence of HHV8 and HIV coinfection in African children in this part of the world. HIV-seropositive children with advanced oral KS have a particularly aggressive course of disease with a poor prognosis [41
]. In sub-Saharan Africa, oral involvement in HIV-seropositive children with KS is common.
Thirty-eight percent (38%) of the patients in this study had common HIV-associated oral diseases which presented concurrently with oral HIV-KS (). It is probable that these diseases were associated with the low CD4+ T-cell counts of the patients, and not with the oral HIV-KS. However, one cannot exclude the possibility that oral HIV-KS-associated cytokine dysregulation may favour the development of other HIV-associated oral lesions and that some HIV-associated oral diseases may further dysregulate the cytokine milieu in the affected oral tissues, thus promoting KS tumourigenesis [42